Remix.run Logo
Imanari 3 days ago

Looks really nice! How does it handle tables?

Adityav369 3 days ago | parent [-]

We have two ingestion pathways: 1. regular OCR + text embeddings; 2. Colpali. We've observed that Colpali does a much better job with tables since it can encode positional stuff and layouts as well.

th0ma5 3 days ago | parent [-]

Whenever I ask people wanting to use such features at scale which figure could be out of place or have a transposed digit it generally makes the project evaporate.